[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y ] [Search | Home]
4Archive logo
So I was talking to a practicing medfag recently...
If images are not shown try to refresh the page. If you like this website, please disable any AdBlock software!

You are currently reading a thread in /sci/ - Science & Math

Thread replies: 92
Thread images: 7
File: Doctor.jpg (23 KB, 266x400) Image search: [iqdb] [SauceNao] [Google]
Doctor.jpg
23 KB, 266x400
So I was talking to a practicing medfag recently about her daily work and I came away with the conclusion that software could easily do most of their work. That does not mean they should be scared losing their work in the near future, because of strict regulations to tackle about responsibility and some people require human interaction, but I feel we are approaching peak MD fast.

It might be a bit counterintuitive, because doctors need lots of high level education and training, and they like to think it is a prestigious job, but her points were the following

1. People can identify common diseases and most can be treated with otc drugs
2. Where symptoms are ambiguous, you need to run some specific tests and evaluate their results
3.a you can either prescribe some drugs to the patient based on some preferences, like medical history etc
3.b send to a specialist for further

And these cover 90% of the use cases. It does not even need a complicated AI but a simple look up table.
>>
>>7801824
>>7801824
absolutely. 200 years from now, if global caliphate or world war III won't happen, people will just step into a box. the box painlessly takes a drop of blood, runs a couple of medical imagining procedures and will suggest a diagnosis/treatment based upon checking thousands of biochemical parameters in a fraction of second. 500 years from now, if ayy lmaos won't wipe out our cosmic dumpster-planet in a preemptive strike, there won't be any illness anymore.
>>
>>7801834
might be, but I'm talking about the next 10-20 years, and that the technology is already here.
>>
>>7801824

Shit, you don't need to do all that. As a medfag myself, if you can just make a program that does two things I'll buy it and install it in my office tomorrow.

> a machine that can diagnose patients with a viral chest infection and convince them that antibiotics won't make them better and them feeling better when they take them is just a coincidence

> a machine that convinces people that they don't need a test if the outcome of the test won't change the management or doing a test without a reasonable Bayesian pre-test probability produces uninterpretable results and useless false positives (e.g. people who get full body MRIs for no reason, screening with PSA, routine chest XR for no reason, etc)

Just doing that would revolutionize medicine and save the system tons of effort and money. Please, please, somebody work on this.
>>
>>7801824
doctor here, I don't think I'm doing magic every day, a lot of the stuff we're doing could probably be automatized BUT I'm pretty sure most people just wouldn't accept to be treated by a machine. Supposedly we're doctors and we've been studying for many years and yet when you've got a patient it can take a lot of psychology and time to get him to trust you and listen to your advices. Even if the perfect program was installed right next to my office tomorrow I'm still sure people would still check with me or another doctor to make sure the machine is right. And here I'm just talking about the very basic viral infections, the more serious disease it'll be and the less likely people will want to put themselves into the hands of a machine.
>>
This sounds like the movie idiocracy where you stick plugs in your mouth and anus and are told what is wrong with you and what you need for whatever is wrong with you. The workers at the hospital are basically McDonald employees who press a button and are straight retarded.
>>
Just give me access to the equipment, and let me pay a deposit. I think self-care is where it's going.
>>
>>7801856
word
>>
>>7802381
This

Also, medicine isn't as simple as people belive. It's not symptom X + symptom Y = diagnosis Z. In many cases, it's hard to even get the patient to specify the symptoms that they are experiencing. It takes a human being, with empathy and ability to understand other people, to really understand what the problem is. I, as a medical student, didn't really understand this until I got to the clinic and met patients for real.

Sure, many primary care cases could probably be automatized. But when you get to the hospital or emergency level, things are a lot more complicated.

Who is going to examine the patients? Who is going to decide what questions to ask? Who is going to operate on the patient?(surgery is a long way from being automatized).

tl;dr medicine is more complicated than one would think. We are a long way from automatization.
>>
>>7802455
>Also, medicine isn't as simple as people belive. It's not symptom X + symptom Y = diagnosis Z

Correct. And machines can handle that non-linearity far better than humans ever will be able to. Diagnosis is a clear cut data science problem, and humans just aren't good at it.

I agree that humans will be needed for the "data collection" part, but doctors will basically just become human interfaces for machines that are managing the treatment plans for patients. The doctor will just be there to sign off for legal purposes, and people people will feel uneasy taking medical advice from machines.
>>
>>7802470
Diagnosing isn't really the difficult part of medicine. Once you have the data from labs and clinical examination, the diagnosis is very straight forward. No software needed.
>>
>>7802473
>implying there is a difficult part to medicine
>>
I don't see the elimination of "doctor" jobs - I think 50 or even 100 years from now, most people will still have a human GP and interact with humans during hospital stays and so on.

What will change is the education required to become a "doctor". I doubt that in 50 years your average GP will have gone through a decade of grad school. It will become essentially a humanities job, where the main task is to provide a human interface for the patient. All the technical stuff will be handled by machines, the job of the "doctor" (a new title will probably be invented) will be to a) get patients to a point where they're comfortable discussing their symptoms and reassure them if necessary and b) take detailed notes that can then be fed to a computer that will provide the diagnosis and treatment recommendation. I see it being a a job that requires a bachelor's in any field (english, psych, social science being preferred) plus an extra year or so of vocational training.

There will still be proper doctors who will be highly educated and highly paid, but patients typically won't interact with them unless something is seriously wrong. Consequently, there will be far fewer, since the "relatively lower-skilled worker plus sophisticated computer" formula will be sufficient for the majority of patients.
>>
>>7801824
It all depends on what particular kind of medicine you're talking about.

>GPs
Yes, get rid of the fuckers. Waste of time and money and less useful than just buying PDR with the exception that they can prescribe drugs, which they inevitably fuck up because they can't keep interactions straight.

>Emergency Room Doctor
These guys stay because you need a level of abstract thought that computers aren't good enough at yet

>research doctors
I'd kinda say get rid of these fuckers too because other disciplines can into research far better than MDs

so on down the line.

We won't get rid of them for a long time though because of feelings and that human touch thing.
>>
>>7802699
The problem with medicine is that you need appropriate experience and knowledge to interpret a symptom. Typically a young doctor will see the same patient in a much different light than an older, experienced doctor. On the other side even a very experienced nurse will be at best able to tell you how she feels about a patient, sometimes even know the basic algorithm to take care of him, but she'll rarely be able to really understand why we're doing what we're doing and therefore she'll be at a loss when you have to react to something that doesn't go as expected.

Even for very frequent diseases like appendicitis, there is no sure way to make a diagnosis. Sometimes every tests come back normal at first and then 24 hours later the person has to be operated for severe peritonitis. Right now medicine isn't advanced enough that you could just code every algorithm into a computer and expect good results.
>>
Yes, yes and yes. Soooo I went to the Optometrist (eye doctor) today. So much needs to change.

The entire process was chaotic and traumatizingly bad. The entire office looked like a riot went through it; everything was tornup and scuffed. Papers scattered EVERYWHERE.

Some unqualified local hick female in a did all of the leg work (worked the refracter and everything, did the puff test) and at the end the doctor showed up and was like let me look in your eyes. Shes all like lmao idgaf your eyes are there; that'll be $160 bucks.

Before I could leave they made me sit down with the most downsy friend they could think of for the job so that she could try to set me up with glasses. I'm like, "Lol, whatever I'll humor this...." Listening to her spiel. Tell her I might be interesting in frameless glasses. Ask her how much with lenses and she pulls out a calculator. She says it'll be $657 for a pair. I contain myself from standing up right there, turning 180, bending over with one leg on the table and blowing shit all over her face. I chuckle and ask how long the turnaround is if I order them today. Two weeks. Say, I can't buy now, need wife's approval.

As soon as I get home. I check online for cheap prescription glasses. Order a pair of custom prescription frameless glasses with a coupon for a grand whopping total of $45 bucks.

All the staff were fat catty women, and one of them had obviously suffered from fetal alcohol syndrome when looking at her nose bridge (wtf, so they appropriately gave her the hardest job in the office of being the financial checkout lady so that she could make my departure as excruciatingly delayed as possible).

I seriously don't know how being an optometrist is a real job.
>>
>>7802381
This is true as of now, but maybe later in the future people will be apathetic towards a doctor's approval or advice.
>>
I mean people get excited by self driving cars, which are cool I guess, but don't really improve anything. But it is a way harder problem to develop a capable AI that you can let roam freely on the streets. Diagnosis are very low hanging compared to that. Like Akinator can guess any character I can think of in 30 questions with pretty vague answers.
As if doctors are just there for controlled drug abuse.

>>7802381
that is a pretty weak reason to justify the costs, especially in poorer countries

>>7802819
If 2 doctors interpret the same symptoms a different way someone is doing a bad job, just saying. Let it be lack of knowledge or inexperience

>>7802925
This, like witch doctors. A proper software could model and examine interaction House not even could think of
>>
>>7802895
You sound like literally the most arrogant, insufferable cunt I have ever met. Your capacity to express such terrible personality traits through writing style alone is honestly breathtaking.
>>
File: belgian_health.jpg (45 KB, 634x423) Image search: [iqdb] [SauceNao] [Google]
belgian_health.jpg
45 KB, 634x423
>>7802895
>one of them had obviously suffered from fetal alcohol syndrome

not her fault though.
A bit orthogonal, my pet peeve is when people give me shit about my cholesterol levels seeing me eating eggs after workout and how too much protein is bad for you. You don't even lift bitch, I read more medical papers about all that stuff in a week you ever will. Like you can cite some bullshit livestrong article and suddenly you are a dietitian.

pic related, Belgian health minister lecturing about bad diets and eating habits
>>
>>7801824
I'm a medfag and I agree there is much to be said about the use of 'expert systems' or software to assist in diagnosis.

That said, they'd just be a tool for people to use. They can't replace us because the patient isn't going to be educated enough to make use of the software or self-diagnose, nor does the computer have the ability to actually treat the patient.

Also, robots have a well-documented and explored trait of having an awful sense of touch. Otherwise robots would be carving up the cattle and pigs and chickens in the slaughterhouses, not 3rd world undocumented migrants. If a robot cannot autonomously and reliably carve a decent tenderloin, they sure as fuck cannot autonomously remove a cancerous tumor near some vital organ on a person.

Maybe in the dystopian libertarian future where things like tort law, medical malpractice and such shit exists, there will be sketchy medical robots in vending booths ready to perform diagnosis and gallbladder surgery, clean up and follow-up all autonomously with the swipe of a credit card. Sadly, this is unlikely to happen.
>>
>>7803210
>pic related, Belgian health minister lecturing about bad diets and eating habits

At least she'll die having lived a fucking life, eating waffles and smoking weed.
>>
>>7803218
that is because you are focusing on the exceptions here. Surgery is a serious deal and much can go wrong. But I'm pretty sure they are much more precise and their 'hand' are very firm, hence teleoperated surgical systems. And why would they touch you when they can operate lasers?
>>
>>7803235
A robot with a laser is going to do even worse than a robot with a scalpel. Lasers cut by BURNING. Why not just reprogram an industrial C-beam plasma cutter from the old Saturn Flint, MI factory for kidney transplants?

Computers are awesome tools, but they're not going to replace anybody in the hospital. Not even the ER janitor or the security guard or even the nurse at the front desk or the emt student that does nothing but take vitals for the head nurse.

It's not even worth it. The EMT/nursing/medical student costs nothing to operate. A robot that does nothing but take vitals and NOT create a lawsuit against the hospital would probably cost a million dollars or more.

If you want some kinda queer OA transhuman future you need to target politics and law.
>>
>>7801824
>can be treated with otc drugs
lol no.

>>7801824
>send to a specialist

A lot of doctors are specialists.

and how do you detemine that algorithmically? Quite often of course a doctor is a retard and a computer would do a better job sure, but how do you figure this out? Strong AI isn't here yet.
>>
>>7803263
what do you think shitposting will look like in a wrong of strong AI
>>
>>7803264
>in a wrong of strong AI
Probably makes as much sense as quoted.
>>
>>7803254
have you ever heard of a laser knife? I know you are just funposting, because the whole surgery thing is a huge strawman here, that is not the point

>>7803263
that is the very use case of otc drugs, not going to your doctor with every flu

specialist are required, yes, but that is the same when your GP says you need to go

I'm not a doctor, just guessing here, but they are too following an algorithm. Based on the symptoms, running some blood/urine test for your levels and come up with a probable root cause. I would like to believe that is so strict there is little room for personal interpretation
>>
>>7803289
if anyone have more knowledge on this. I know there is a real shortage of anesthesiologists and they do various tasks. But on a very simple level it sounds easy to give you drugs based on your phyisical/medical records and keep you sleep during operation. You just have to monitor a few signals, it is a simple feedback loop with a very error prone human component
>>
>>7802895
An optometrist is not an eye doctor. Far from it in fact.
>>
wanna become a radiologist

am I fucked?
>>
>>7803298
It's a little different though. Anaesthetics is a discipline where things can go really wrong really quickly because of some fairly small problems, but more than that there's simply a human element involved. We don't really understand the drugs we use for it, so most things are just titrated to effect with some guideline dosages. Side-effects also have to be taken into consideration, dosages adjusted often, different drug choices depending on circumstance etc (in a field where people can't even decide on the better out of two muscle relaxants). It's very much a response-based discipline and we can't replace them with machines yet since we don't know the ins and outs; and even if we did, it's hard to remove the human element from a field with such large variation.
>>
>>7803289
Making a robot which can vacuum and one that can conduct surgery is a very different thing. In one case a malfunction can result in a temporary setback in the other you can have a patient who is doing anything from bleeding to death to having erectile dysfunction and paralysis all the way down their body because a slender nerve just happened to get in the way of te operation.

Another thing to account for is the fact that there are billions of possible levels of anatomical variation that a surgeon has to account from from bones and to muscles to a different layout of blood vessels (for example transposition of the great vessels). Oftentimes a surgeon can't even physically see where they are cutting and have to possess enough experience to be able to predict what structure comes next - be it an artery or a vein. Within that sort of calculation you would have to account for everything from the thickness of the pericardium to the velocity of blood and vessel thickness (to determine what vessel it is although in some places like close to the lungs the veins and the arteries are difficult to tell apart because the pressures are very similar), to rapidity where the operation needs to be performed to the 3 dimension space in which the operation is taking place. And the ever frequent anatomical variation.

I think we are more likely to see /m/ech aided diagnosis first. Right now even the tools used are relatively primitive, I can't imagine those obsessively patient safety oriented organisations approving anything useful in under 10 years.
>>
>>7803949
absolutely, fully autonomous surgery is far away, like self driving cars, fiction. But I believe it will be more like radiosurgery than cutting people with a scalpel.
>>
>>7803218
>medical malpractice and such shit exists
Fuck off you faggot. Seriously. There are so many damn hatchet retard doctors out there that if anything the fucking requirements for malpractice should be lowered.

>inb4 insurance and getting sued are the only reasons medical care is expensive
Fuck you over this too. Medical professionals have been active participants in fleecing patients for longer than they've been considered professionals.
>>
>>7803289
>I'm not a doctor, just guessing here, but they are too following an algorithm. Based on the symptoms, running some blood/urine test for your levels and come up with a probable root cause. I would like to believe that is so strict there is little room for personal interpretation
Fun fact, in the US, they're going by a paint-by-numbers system implemented by whatever insurance company is involved. They cannot deviate or they won't get paid no matter what they think, believe, or to some extent can prove. For the vast majority of common ailments having a doctor involved is just an excuse to jack the costs up. Nurses and CNAs can do basically all of the data collection needed and forward it to the labs (like they're going to do anyway with the doctor around).

This is partly why you're now seeing nurse practitioners growing.
>>
>>7804138
Why do you think self-driving cars are far away? Unlike autonomous surgery, there are already working prototypes for self-driving cars, albeit they only work under perfect conditions.
>>
>>7802347
The problem with your first point is that we as patients don't trust you know the difference between a viral and bacterial infection. We think your too busy counting money, and trying to fly through the latest issue of car and Driver so you can have it in your waiting room. From firsthand experience, Dr's are snobs. Not compassionate caring humans here to help (those are the exception not the rule). I had an orthopedic surgeon flat out tell me he went to harvard, and he runs uofc. So therefore the Dr that referred me to him didn't know what he was talking about. Long story short, the 1st doc did know, and my lawyer called Mr uofc a elitist snob in the court filing.

Tldr; don't lose your humanity climbing the ivy tower of medfaggotry
>>
>>7804327
you answered your own question, and partly because by far I mean we do not have the technology right know, unlike diagnosing diseases. Also, I use this analogy because they both will require enormous amount of legislation.
I would wager, we will have lithotripsy machines autonomously breaking stones before self driving cars on the streets.

Side note, self driving cars are stupid, no purpose whatsoever besides look how geek I'm. Maybe some niche reducing cost in goods transportation if you want to be generous.
>>
>>7804363
>climbing the ivy tower of medfaggotry

exactly the starting point of this thread. Doctors are very smart people and they feel entitled about it. Yet most do a very low tier job, no better than administration and here's the contradiction as people treat them as such.
>>
>>7804651
>Doctors are very smart people
Not necessarily. Doctors worked hard in school to memorize a lot of shit for exams. There's nothing intrinsically complicated or difficult about the vast majority of medical practice, or even med school once you get past the sheer amount of information.
>>
>>7804309
In the 2nd world I usually face that doctors are surprised when I'm asking questions about how to prevent diseases happening to me next time not just taking their prescriptions and leave. They seriously have no idea sometimes, or less informed than someone just reading up on wikipedia and following the links to journals.
Don't really want to imaging what happens in less developed parts.
Now when you have a software it can give you the same quality service without having to rely on your human resources and medical education.
>>
I'm in medschool in Germany, so no worries about losing my job in the future. The ruling party is conservative as fuck, Germany is over-regulated and the weak-minded German population always screams in anxiety whenever an innovation comes around which is ground-breaking (e.g. GMOs)
>>
How can a machine interact with patients successfully, collect subjective information, and understand body language? Furthermore, how can a machine relate to sick humans?
>>7804327
Also, I'm sorry for some of the experience YOU had with doctors being careless and arrogant assholes. That simply isn't the case with all so you need to stop relying on anecdotes.
>>
>>7801824
As someone who has a relatively rare skin disease which only got diagnosed after I mentioned the specific name of the disease to my doctor after reading about it on the internet, I do agree that online databases should and could play a much bigger role for general practitioners, but on the other hand, doctors also usually have a lot of expert knowledge and experience that makes them valuable.
>>
>>7806392
>How can a machine interact with patients successfully, collect subjective information, and understand body language? Furthermore, how can a machine relate to sick humans?

The idea is that while diagnosing diseases and prescribing the right medicine is knowledge intensive, it usually requires zero (0) creativity, which is kind of a straightforward problem where computers excel and can reduce costs.

Handling fuzzy inputs is not that hard a problem and can be eliminated by objective tests (what doctors do anyway). Sure, you can think about ways how you can abuse the system, lying about your condition to get some specific drugs described, but that is not the general point here.

You interact with machines everyday, I don't understand what's the problem here, and I'm sure there will be lots of people offering services where you can pay for their sympathy, while others can get access to better diagnoses than what's available for them currently.

>>7806456
>expert knowledge and experience
That's a thing that computers good at emulating. They are not good at nursing and elderly care for example.
>>
File: DSCF1366.0.jpg (248 KB, 2040x1360) Image search: [iqdb] [SauceNao] [Google]
DSCF1366.0.jpg
248 KB, 2040x1360
>>7806793
>They are not good at nursing and elderly care for example.
Better than some orderlies already.
>>
>>7806793
Objectivity isn't the only part of medical diagnosis and care.People vary and the numbers aren't always indicate what you've read in the books. That's residency 101. Memorizing numbers and symptoms will only take you so far and that's when expertise and experience comes in.
>>
>>7803205
you must be new here
>>
>>7801824

>It does not even need a complicated AI but a simple look up table.

These look-up tables exist already in the form of little diagnostic books which layout diseases and treatments. I have a MD friend who told me that those lookup books are pretty much 90% of a doctors profession, although i suspect in the same sense that Stack Overflow is 90% of being a Software Engineer
>>
>>7804634
>Side note, self driving cars are stupid, no purpose whatsoever besides look how geek I'm

>Less pollution
>Almost zero accidents due driver's fault
>Correct and immediate reaction to most emergencies
>Optimized traffic
>Literally free time for the driver

Yeah, no biggie.
>>
File: hqdefault.jpg (20 KB, 480x360) Image search: [iqdb] [SauceNao] [Google]
hqdefault.jpg
20 KB, 480x360
>>7802895
>Optometrist
>Doctor
>>
>>7801824
How can a computer explore
mental illness?
How can a computer do a
clinic history?
How can a computer know the patient is concealing information?
How can a computer take into account social aspects of the patient?
>>
>>7808270
>How can a computer explore
>mental illness?
GPs aren't qualified to do this.
>How can a computer do a
>clinic history?
Rather easily, especially going forward as more and more medical records are digitized. The beautiful thing is that you can still obey patient confidentiality by using a computer to research parents', grandparents', and siblings' records.
>How can a computer know the patient is concealing information?
Voice stress.
>How can a computer take into account social aspects of the patient?
>implying that most doctors spend enough time with patients to do this.
>>
>>7808279
Your answers tell me that you dont know some aspects of medical practice (GPs can detect and explore mental illness) or that you are biased by your experience with physicians.
Where country are you from?


And other thing that strikes me is that if a computer can know when a person is concealing information why not use it in order to solve crimes.
>>
To suggest that the work of doctors could be replaced by AI represents a severe lack of understanding how western medicine works, and of the social needs of humans. Doctors, of all specialties, learn how to care for patients emotionally, they learn how to effectively gather information about causes and symptoms, they comfort family-members, and so much more.

Even acting simply as a diagnostician requires a human brain. Very few symptoms are caused by one disease, and diseases can present radically different symptoms in different people. A guess-and-check algorithm will get you only so far.
>>
>>7808300
>GPs can detect and explore mental illness
GPs are legally allowed to. They are by no means qualified to do so though.
>>
>>7808340
>To suggest that the work of doctors could be replaced by AI represents a severe lack of understanding how western medicine works, and of the social needs of humans. Doctors, of all specialties, learn how to care for patients emotionally, they learn how to effectively gather information about causes and symptoms, they comfort family-members, and so much more

http://www.alexa.com/siteinfo/webmd.com
WebMD is the 99th most popular website in the US.

People *hate* going to the doctor.
>>
>>7808350
Right. So when anxiety about an ailment you're experiencing isn't quelled by WebMD (or in the old days, a book), you go to a doctor. Where they care for you, work with you, and bring you comfort and a treatment protocol. Also WebMD alone should prove my point. Put in any common symptoms and read the thousands of search responses.
>>
>>7802381
>I'm pretty sure most people just wouldn't accept to be treated by a machine.
They wouldn't be "treated by a machine". Most of them would be treated by a person using a machine, a trained operator.

We should have switched over to this for most routine cases at least twenty years ago, but doctors have been the ones holding this up, digging in their heels and fighting hard against anything that would reduce the demand for their services, just as they've fought hard against training an adequate supply of doctors to meet demand, because you scum value your own wallets over human life so much that you've been pursuing a policy of total employment through engineered shortage.

>Supposedly we're doctors and we've been studying for many years and yet when you've got a patient it can take a lot of psychology and time to get him to trust you and listen to your advices.
Doctors are wrong all the fucking time, from being ignorant, from being lazy, and from acting confident when they're just guessing, and are rarely stripped of their license to practice for anything short of being caught raping or deliberately murdering a patient.
>>
>>7808340
>Even acting simply as a diagnostician requires a human brain. Very few symptoms are caused by one disease, and diseases can present radically different symptoms in different people. A guess-and-check algorithm will get you only so far.
Expert systems have been outperforming human doctors at diagnosis since the 1970s. "Guess-and-check" is exactly what doctors do, and which comprehensive diagnostic algorithms avoid.

What you do is start with symptoms, and the system comes back with questions and orders for tests. It will keep track of and look at ALL available data, not just let its attention wander through and hopefully catch on something that matters.

Diagnosis of known diseases is a lot like playing chess: humans can be pretty good at it, but once you program a computer to beat the best human player, you can copy it around the world and it will go on playing to the same world-class standard every time, without regard to fatigue or mood, forever.

We will continue to need medical researchers, special experts to handle cases too unusual or new to have been incorporated into the algorithm, and procedure-oriented practitioners such as surgeons, but not workaday diagnose-and-prescribe medical doctors.
>>
>>7803210
>You don't even lift bitch
most people who do lift have dumb bro science tier ideas about health though
>>
>>7808350
It might also be that people are afraid of going to the doctor.
What´s more, even if your correlation were true, your argument only can be applied in US (where there are serious problems in health policy)

>>7808343
And you know what they need to be qualified?
Not all mental illness are schizophrenia

>>7808505
What can I say? With your great "ad hominem" without proof, you have convinced me completely.

>>7808543
>Expert systems have been outperforming human doctors at diagnosis since the 1970s
[Citation needed]
>>
>>7809383
>And you know what they need to be qualified?
>Not all mental illness are schizophrenia
More than a single seminar/class on mental illnesses. Think about it for a second, a GP is legally allowed to treat the same disorders and prescribe the same drugs that a psychiatrist goes through a 2-3 year specialization for, and psychologists spend undergrad and graduate school learning to treat.

This is one of the roots of the fucked up mental health system in the US. We have too many idiots who have no clue what they're doing like GPs and social workers given Dx power in a field that they have no respect for and no real knowledge of.
>>
>>7804363
What a compelling story

Next time, have a point
>>
>>7809579
I don´t know how it is in your country but in my career we have "more than a single seminar/class".

Obviously you are right and diagnosis/treatment of mental illness needs to be done by a psychiatrist but I was saying that a GPs can explore mental state of a patient (through clinical interview) and referral to a specialist.
>>
>>7809637
In the US it's usually no more than around two classes, if that. It's touched on in a couple of classes in med school, but there's no real depth to it.

Further in the US, they often do not refer to qualified professionals for mental health issues (admittedly this is often for a variety of reasons). They diagnose and then treat (usually with medication) most of the more common disorders (the various flavors of depression, anxiety disorders, ADD/ADHD, etc.) quite often getting them entirely wrong while never addressing environmental issues.
>>
>>7802347
I really feel for you, doctor-bros.

I have something wrong in my ear canal. Thought it was a pimple until it got worse. Thought it was a viral infection until it didn't go away for a week, started to get worse.
This has happened before, so I know the general ins and outs of ear stuff, minus stupid complications.
Found an expired z-pack at home, read up that medications are actually still good after expiration thanks to military testing the stuff (there's a harvard paper I can link if anyone cares), started the anti-biotics a few days ago, getting better.

I dislike how people go to the doctor for a runny nose or cold, or how people decide that the emergency room is the first choice for everything.

/rant
>>
>>7803263
protip, fellow american.
In other countries, anti-biotics can be bought over the counter.
Here too, if you don't mind taking "pet" antibiotics. Same stuff, just can't be legally marketed to humans.
>>
>>7809579
Honestly half of the problem are psychiatrists and general practitioners throwing drugs at everyone and everything.
A good percentage is the american/human problem of just wanting a quick fix, but still.

t. angry psychologist who was also forced to take ADHD medication despite not having at as kid.
>>
>itt butthurt stemfags that couldnt get into med
>>
1/2

>>7809771
I'd assume the expiration date depends on the substance in question. I would be interested in the study.

>>7809782
I'd be careful of generalizing too much. Many countries in Europe decided that creating thousands upon thousands of new, antibiotic resistant microbial strains (especially bacteria) each year wasn't such a good idea after all, so you'll have to have a doctor's slip for most pharmaceuticals above paracetamol in many countries.

Taking a look at MRSA resistance in the world paints a bleak picture (pic related). My country (Norway) is pretty much the only country along with Scandinavia that isn't infected beyond repair, though that will inevitably change. You can't even work in a hospital here if you are infected: You'll need to undergo a strict cleansing regimen for a long time (months), often several times, and hope that you can rid yourself of the strain in question.

If you are a foreigner here, you shouldn't (but often will be due to no enforcement and lacking routines and training) even be admitted to the hospital; the rule is that you will have to wait outside of the hospital in question and change into hazmat gear before being allowed into a quarantined room if there is any suspicion of resistant strains (which there will be if you are from anywhere but Scandinavia).

For those of you that don't know much about medicine, the problem with MRSA and other resistant bacteria isn't that they're particularly dangerous, it's that hospitals house a lot of sick people, many of whom have less than stellar immune systems. They might catch a resistant strain and that strain might kill them because their immune systems aren't fighting the invasion, and you can't combat the bacteria with antibiotics because the bacteria are resistant. Most of the time (but not always) there are other antibiotics that also work against the strain in question, but we're quickly running out of pharmaceuticals.
>>
2/2

>>7809909

In the end, we can thank retarded doctors and retarded patients for this. Oh, and Spain. Especially Spain. Fuck you, Spain.

If you start some antibiotic treatment, it is of utmost importance that you finish it as prescribed (i.e. take it for e.g. 20 days even if you feel fine after 5 days), or you risk mutating a bacterial strain into being resistant.
>>
>>7809910
Here in Canada MRSA (and VRE, etc) colonization is still the exception rather than the rule but it's certainly around - so we do swabs on admission to hospital, if positive isolate and gown and glove for all patient contacts.

The other problem with MRSA is that if it is prevalent in the population, it changes the first choice antibiotics for severe infections of unknown origin to be more expensive with (maybe) more potential side effects
>>
>>7808221
that is called a train/tram, cars in cities should be banned already. But let's just throw endless money at a problem with good alternatives.

>>7808367
>So when anxiety about an ailment you're experiencing isn't quelled by WebMD (or in the old days, a book), you go to a doctor.

because she is the only one that can prescribe drugs that actually work?

>>7808543
word
>>
>>7808270
>How can a computer explore
>mental illness?

I don't even know how good humans are at mental illnesses? According to what I read on the internet a large portion of Americans have some kind of drugs described for mental disorders. Yet I never met anyone in my country treated for ADHD. I'm not even sure it is considered a disorder here.
Some who specializes in it and spends her days treating them sure, but your average GP? Nah.

Or you can think of it this way. If we have computers picking the low hanging fruits, we can have more mental illness specialists.
>>
>>7808270
>How can a computer know the patient is concealing information?

Putting aside all the NSA shit, you have no reason to lie to a computer when there is no repercussion for it. Confidentiality is a strong point for E doctors imho.

>How can a computer take into account social aspects of the patient?

for what?
>>
Really? This far and no one has posted this?

Oh, wait, we're all too cool for xkcd because it's only for tryhard wannabes or some bullshit.
>>
As long as most healthcare shit is paid for by government, you can't expect any reasonable cost control.
>>
>>7811306
Yes, the free market has done wonders for lowering the cost of privatized healthcare in America. It's a wonder more countries don't opt for it.
>>
>>7808350
People hate going to the doctors because they will jew them thousands of dollars.
>>
>>7811366
Actually, most everything that has inflated the cost has been government interventions and not free market forces.
>>
>>7802455
Except for surgical procedures, everything else you've said can be currently automated. Going in the medical profession myself, I actually come from a computational background and I plan on trying to help bring more automation to the medical field.

Its appalling to me that medical charts are still localized and on paper in some hospitals.

As for "empathy", have you considered that the human factor actually hurts diagnosis, since people are often embarrassed or afraid to say their symptoms, or because there is a language barrier or many other "human factors".

In reality, we can lower costs and increase efficiency by replacing diagnostician with a technician.
>>
>>7811893

>!experience ++opinions

you really think you're the only tech savvy person (not even in yet) in medicine?
>>
>>7810966
>ITT People that don´t understand medicine trying to talk about medicine
>>
>>7802426
>$400k
>highest paying job in the world

Try again dipshit
>>
>>7812508
its claiming the pay/effort ratio is the highest in the world.
>>
>>7812393
>muh job

;^)
>>
>>7812672
>Muh I can talk about things I don´t know
(^;
>>
>>7812508
I'm curious, what is the highest paying job? I mean net salary, no bonuses, and just running your company, but a real employee?
>>
Itt: Geeks talking about something they don't know about. Yes, I'm a medfag. I don't write about space or something I don't know because I know my opinion doesn't have solid bases cause that is not my field. Same with you.

>hurrr medicine is just an algorithm

You have a "point" here, we have some medical guides that even include some algorithms. But if you expect a computer to do this you will only give a symptomatic treatment, but you won't resolve the main problem.

>just machines with all the medical information will do the job

Well, do you remember that Shit on your skin that you Googled and it came out it was fucking cancer? I see this Shit everyday, patients come to us with a fucking diagnostic made in Google, and when you give them an explanation they don't believe you. They eat antibiotics like candy and when you give them the right one, Shit doesn't work because they already developed some resistance. WTF people. Believe in us, fucking 10 years studying this Shit and you still believing in your grandpa's solution.

I think the main problem is the healt system, we have familiar medics, they care about you even when you are not sick, they help you to prevent Shit, but you don't even know them. People just go to the medic when they are sick. And when the main problem is too big to being resolved with antibiotics, they throw Shit at you.
>>
>>7813442
>Yes, I'm a medfag. I don't write about software engineering or computer science

ftfy
Thread replies: 92
Thread images: 7
Thread DB ID: 470890



[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y] [Search | Home]

[Boards: 3 / a / aco / adv / an / asp / b / biz / c / cgl / ck / cm / co / d / diy / e / fa / fit / g / gd / gif / h / hc / his / hm / hr / i / ic / int / jp / k / lgbt / lit / m / mlp / mu / n / news / o / out / p / po / pol / qa / qst / r / r9k / s / s4s / sci / soc / sp / t / tg / toy / trash / trv / tv / u / v / vg / vip /vp / vr / w / wg / wsg / wsr / x / y] [Search | Home]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.
This is a 4chan archive - all of the shown content originated from that site. This means that 4Archive shows their content, archived. If you need information for a Poster - contact them.
If a post contains personal/copyrighted/illegal content, then use the post's [Report] link! If a post is not removed within 24h contact me at [email protected] with the post's information.